If you were waiting for a new Renaissance of software development and super-networked self-starting educational bootstrapping to spread across the Third World as a result of the OLPC program... well, now could be a good time to find yourself some beer, and then start crying into it.
Parts of what I wrote in One Laptop Per Me are still perfectly applicable. The other parts, in which I just sort of assumed that the OLPC people wouldn't turn the whole deal into one giant WTF, now seem less well-chosen.
Oh well.
If, on the other hand, you're a cheap-tech vulture who just wants to suck up an OLPC laptop or three for fifty bucks each, the apparent utter debacle that the OLPC program seems doomed to become can only mean a buggerload of shiny white-and-green mini-laptops will be swamping eBay any time now.
(No, they never did turn out to make different models for poor kids and rich Westerners. So if you buy an eBay OLPC that was stolen by the random dude chosen to take a truckload of laptops to some not-even-served-by-the-Post-Office outpost in Peru, nobody will be able to tell that your laptop was originally meant to be given to a poor little kid. You may still feel guilty about buying it, and I hope you do, but if all this is true, then it seems depressingly clear that no force on earth ever actually would have gotten that computer into the hands of its intended recipient. Some bastard's going to buy it on eBay. Might as well be you, I suppose.)
The above-linked essay by Ivan Krstic is just one ex-OLPC-employee ranting, so everything in it may be nonsense.
But unless the core claim in that essay - that OLPC is basically completely without a deployment department, so there's nobody to make sure the half-million laptops they've sent out actually get where they're meant to go and can be made to work when they get there - is fundamentally false, then pretty much the whole OLPC project is, as of now, dead as a stone.
(Oh, and you know that special security system that's meant to disable stolen XO-1s? it turns out that Ivan Krstic was the main architect for that system, yet does not mention it in his essay. Which suggests that he's not hopeful that it'll do anything in particular to actually prevent theft.)
And this isn't even mentioning the earlier problem, that the constructionist philosophy that the whole OLPC project is built around has never actually been shown to work on any significant scale. Constructionism sounds as if it ought to work, but nobody's done it yet.
(There's also some closed-source versus Free Software folderol, which in this case does have a bit more bite than usual, if only because the XO-1 laptop has a "View Source" button that's supposed to show you the source code for pretty much anything you're looking at. This makes the concept of a Windows XP XO-1 particularly poignant, but I agree with Krstic that this is hardly the central problem, particularly seeing as the XO-1's "Sugar" GUI is what the View Source button is meant to affect, and Sugar can unquestionably be made to run on Windows, somewhat like unto Windows 95 on DOS.)
Krstic goes on to describe OLPC as an impending "historical fuckup unparalleled in scale", which is a great exaggeration. If OLPC turns out to be an utter and unmitigated failure then I suppose it might perhaps just make it into the Top Twenty of information technology fuckups (here, from just the other day...), but I suspect it won't even be in the top couple of hundred of historical business debacles, let alone certain military fuckups I could name.
And all this doesn't, of course, mean that the computers-for-the-poor idea is forever doomed. There's a whole new wave of low-cost mini-laptops, headed by the Asus Eee PC, which was pretty much kicked off by the OLPC XO-1. I think it's quite likely that these new systems will leak out into the Third World in a year or three, when things like the original seven-inch Eee start to become bargain-basement items.
It is amazing what kids - even very young kids - can do with computers, whether or not those computers arrive covered with spiffy unproven constructivist gift-wrap. And wireless-enabled solid-state low-power-consumption systems suitable for use out in the boondocks will become cheaper and cheaper as the years go by.
And heck, OLPC isn't targeting only poorer nations; there's an independent OLPC office here in Australia, for instance.
But you can't expect even the most enterprising of schoolchildren to pull laptop charging stations out of their fundaments, figure out programming from first principles all by themselves, or go and catch the bad man who drove all of their laptops over the border and swapped 'em for a truckload of dope and vodka.
Can it really be true that OLPC just ignored these issues?!
(Slashdot discussion of the essay here. Much Free/Non-Free Software heat, not much light. Expect a gun-control thread to start about half-way down the page.)
15 May 2008 at 3:59 am
"You have to run with the kite."
"Nonsense. It's supposed to fly."
15 May 2008 at 10:39 am
"Phase 1 = Collect underpants."
"Phase 2 = ………"
"Phase 3 = Profit!"
Or in this case - happy, educated, tech-savvy, third-world child programmers!
15 May 2008 at 2:52 pm
What if it worked?
Somewhere in Democratic Republic of Armpit:
10:00am - truck with laptops makes it through to village.
10:15am - laptops distributed
10.16am - first laptop screens broken in sibling fights
10.23am - first laptop bios corrupted
11.47am - all laptops in local wifi-mesh permanently destroyed by malware downloaded from warez/pr0n site.
11.48am - Game over, man, GAME OVER!
15 May 2008 at 2:55 pm
Having said all that, I did go and buy an EEEPC, and it's absolutely brilliant - exactly what I have always wanted from a laptop (small, light, simple, typeable, robust), and also why it took me to now to buy one. Lucky I've got small hands though!
16 May 2008 at 3:10 am
maybe olpc is a colossal waste in terms of education returned on money invested, but i still think the project is a success because:
- it certainly will get *some* laptops into the hands of children
- there is the hope that those laptops will 'prime the pump' of information flow into developing communities.
i question the real need for a deployment department. i think that maybe it's more important to build vast numbers of cheap laptops and sprinkle them around the globe like faerie dust than it is to carefully ensure the proper operation of each delivered laptop. the shotgun approach will get them out there, and then if they're interconnected well enough they will facilitate the spread of knowledge. damn the man, save the empire. i know if i grew up some dirt-eating somalian or whatever i wouldn't give a shit whether or not my olpc came with deployment support or not. i'd just need it to work at least a little bit. the rest will follow. people are capable.
16 May 2008 at 6:37 am
I think you misunderstand what deployment support means nmr8. It's not about making sure the laptops that get delivered work OK... it's about making sure the laptops get delivered at all. It's proving to be the exception to the rule when a laptop actually makes it into the hands of the intended recipient. More often than not the laptops are being stolen and resold en-route. This is due to a wide variety of reasons but a big one would lack of planning on how to actually get the laptops to remote sites - which is what deployment support actually is. It's probably better termed as logistical support.
16 May 2008 at 10:42 am
The logistics of foreign aid is the furthest thing from the mind of the person dropping coins into a 'save the kiddies' bucket.
Sorry to be rather pessimistic about it all, but I've been out with an aid officer working in the forests on the Thai-Burmese border. The work was community and commercial development; specifically helping local women set up their own businesses. Not to trade overseas and produce 'ethnic' trinkets for westerners, but simply trade within their region. Basically a one-person TAFE & business college. It was scary how little actual, tangible "aid" filtered through; at every level someone took a cut, and this was actually in an organisation that has a deployment division! The most heartbreaking, soul-crushing depressing incident was Christmas - presents collected by kids back home went to kids of embassy /aid office staff, not kids out in the jungle. Their rationale was "Thais don't celebrate Christmas anyway".
A question - who did the localization at OLPC? Who was their linguistic expert for Uganda/Peru/Upper Nowhere? I hope they remembered that little step!
16 May 2008 at 11:22 am
If someone was to offer me a free OLPC, I'd take it- but it's not what I'm looking for in a laptop. Neither is the Eee PC, although it is closer.
(Sigh) If only there was a company making custom laptops. There are plenty that allow minor modifications to their basic design- but not start from scratch.
So, no 11" touch screen machines, with IBM Thinkpad-style butterfly keyboards. Still- a man can dream, I suppose.
16 May 2008 at 3:29 pm
You guys seem like a smart bunch - how could you make a security program/device that only allows a child to operate the laptop?
A determined thief can steal anything, but my rationale here is that any sort of deterrent will cut out a large amount of theft. It could be a false premise but I'm interested to know if you guys can think of anything...
For example - how about a power switch fingerprint scanner that only looks at the size of the finger? It wouldn't work for older children and some people with small hands could operate it, but ces't la vie. Or can you determine the age of someone from their retina?
17 May 2008 at 4:32 am
Biometrics are a singularly unsuccessful way of doing verification of even the identity, let alone age, of a user. There are some newer systems that seem to work better than the older finger/eye systems but they would not fit into a laptop - specifically hand blood vessel mapping using a near infrared scanner. http://www.fujitsu.com/downloads/COMP/fcpa/biometrics/palmsecure_datasheet.pdf . Too bad as well because it would not be too tough to add an algorithm looking at hand size to determine a rough age grouping for the user.
Really though, there just isn't a particularly good way to do this. One idea though would be to geographically limit the operation of the units - maybe via GPS - so at least units originally destined for a specific area of the Congo won't work in San Francisco (or anywhere else for that matter). Or you could simply create RFID security tags keyed to individual laptops that the laptop must be within range of to boot. Deliver these tags completely separate of the laptops. These ideas have some obvious flaws though and, like most security measures, wouldn't stand up to a determined and even moderately skilled hacker with physical access to the unit. Not to mention that you are still stuck with the same initial problem - no assurance that the items ever get to the intended destination. The best fix for this would be establish a deployment group and actually give them enough money to get the items where they need to be - of course this would be expensive and probably push the price of the OLPC up quite a bit... which kind of defeats the purpose.
21 May 2008 at 6:31 pm
"maybe olpc is a colossal waste in terms of education returned on money invested, but i still think the project is a success because:
- it certainly will get *some* laptops into the hands of children
- there is the hope that those laptops will ‘prime the pump’ of information flow into developing communities."
i dont think it works. a few laptops doesn't mean much if the resources were taken from the same pool that could have been used to pay teachers and build infrastructure. if the olpcs were all donated sure, but they aren't.
as for the flow of information, look at the middle east, they get satellite tv and internet, and nutty beliefs and conspiracy theories are still rampant.
mainly, the widespread access to pc's in western countries has not created a flood of super literate math genius's. the magic bullet view of technology is incredibly flawed, and that they can't see the evidence just shows how naive or ideologically blind the olpc people are.